GB2527097A

GB2527097A - Apparatus for non-volatile indication of a fault on a data storage system

Info

Publication number: GB2527097A
Application number: GB1410434.3A
Authority: GB
Inventors: Alexander Carl Worrall; David Michael Davis; Odie Killen
Original assignee: Xyratex Technology Ltd
Current assignee: Seagate Systems UK Ltd
Priority date: 2014-06-11
Filing date: 2014-06-11
Publication date: 2015-12-16
Also published as: GB201410434D0; US9779006B2; US20150363253A1

Abstract

Disclosed is a storage system with a plurality of storage media and means 18 of indicating fault information relating to a component of the storage system. The means include a display arranged to display the fault information and configured such that when power to the apparatus is cut-off the fault information remains displayed. The indicator means may mounted on a carrier 12 that can be removed from the storage system. The display may be an electrophoretic or e-paper display. The display may show alphanumeric data or be arranged to show the relative position of their associated components.

Description

Apparatus for Non-Volatile Indication of a Fault on a Data Storage System The present invention relates to a storage system comprising a plurality of storage media and an apparatus for indicating fault information relating to a component of the storage system and to a method for indicating fault information relating to a component of a data storage system.

As the complexity of the computer systems demanded by research and industry increases so too does the amount of memory required to support such systems. The growing number of components involved in large-scale data storage makes diagnosing and repairing problems progressively more difficult. It is critical for detailed information about faults to be easily accessible so that when something does go wrong the problem can be pinpointed and dealt with.

Large-scale data storage systems generally comprise storage devices such as solid state or hard disk drives held in carriers which are then slotted into much larger enclosures or drawers, themselves mounted on racks. This way a high density of data storage devices can be achieved. An example of an enclosure of this type is shown in figure 1. SSDs are mounted in carriers 6 which slot vertically and side by side into enclosure 4. It will be understood that adaptions to systems are ongoing and that similar systems may, for example, contain carriers or components that are differently orientated or differently sized. These storage systems also generally include power modules (PSU5), electronics modules with controllers providing input/output connections to the enclosure and carriers for cooling the system which are usually removable and replaceable.

The failure of an entity (for example a controller or other device) can be communicated to the user by means of an indicator such as an LED that is mounted to the enclosure so that it is visible externally during operation. The problem with such a setup is that these entities are usually complex and may comprise multiple sub-modules or components; for example, a controller may contain several DIMM memory slots or a storage carrier may house a number of SSDs. In most cases where many components are present and one of a number fails the failed component will not have its own

I

indicator. The user will be notified that the carrier containing the component has a fault.

Once it has been established that a particular carrier has a fault it must be removed, at which point power to it is lost and the fault indicator or indicators cease to function.

Separate components may then need to be tested in order to establish precisely where the fault is located before steps can be taken to replace the component or repair the fau It.

According to a first aspect of the present invention, there is provided a storage system, comprising: a plurality of storage media; and an apparatus for indicating fault information relating to a component of the storage system, the apparatus comprising a display arranged to display the fault information and configured such that when power to the apparatus is cut-off the fault information remains displayed.

The invention provides a simple, compact way of indicating the status of components within a carrier after the carrier has been removed from an enclosure.

Advantageously, the fault information will persist for a significant length of time after disconnection of the carrier meaning that a user can set it aside and at a later time, or on inspection by another user, it can quickly be determined which component within the carrier needs attention.

Thus, a means is provided by which information about faulty components inside a carrier can be recorded and retained for an extended period after the carrier has been removed from the storage system.

In an embodiment, the apparatus is mounted on a carrier that can be disconnected and removed from the storage system. As described above, storage systems generally comprise carriers holding modules for storage, cooling, electronics etc. Being able to disconnect these carriers and extract them makes repair work much easier. The advantage of the non-volatile display screens for fault indication is that when the carrier is disconnected in order to access the components within, the information required to fix the fault is still visible. The components in question can be any type of components associated with storage systems, for example SSDs or fans for cooling.

Coupling the apparatus to the carrier means that when the user begins a repair job on a carrier that has been extracted from a storage system, information regarding the faulty

I

components is readily available. It is possible to mount the display elements on the exterior of the enclosure or on some other part of the system which will allow them to be viewed when the corresponding carrier is connected up, however in such a case it is then necessary during repair to return to the enclosure to determine which components need attention.

Preferably, the display is an electrophoretic display. Electrophoretic display technologies, which in their simplest form involve controlling the appearance of a screen by moving charged pigments in an electric field, are rapidly developing and are integral to devices such as electronic readers (e.g. Kindles®) or mobile phones. Light is reflected rather than absorbed by materials in the display which mimics the appearance of paper much better than an emitting screen.

Earlier versions of the electrophoretic display work by sandwiching a solution containing small, charged, titanium dioxide particles suspended in hydrocarbon oil and black pigment between two electrodes, the upper (viewing) side of which is transparent.

Additional substances are added in order to prevent the particles from losing their shape and cause them to become negatively charged. When an electric field is induced between the electrodes, with the upper electrode acting as positive and the lower electrode acting as negative, the white negatively charged titania particles move towards the positive electrode at the transparent upper surface and cause the area in question to appear white. When the current is reversed, particles move to the opposite pole and the pigmented oil becomes visible at the upper surface causing the area to appear black.

The surface of the electrophoretic film can be divided into pixels in which an electric field can be independently controlled. E Ink Corporation developed a similar product using microcapsules to contain the oil and titania particles. This allowed a flexible display to be produced and helped to minimise unwanted effects (such as horizontal movement of the particles). For a detailed description of such a display see US-A-6,515,649 (E Ink Corporation) or US-A-5,930,026 (MIT).

Other more recently developed versions of the technology include electrowetting' displays (controlling the behaviour of a film of coloured oil) and eletrofluidic' displays (changing configuration of a coloured pigment from a small reservoir to a visible layer at the viewing surface). Development in multicolour displays is also ongoing.

According to some embodiments of the present invention the electrophoretic display is coloured (via any of the methods known in the art such as use of filters or an electrofluidic display using three or more coloured pigments) and/or may bend over edges of a carrier. The cost of these displays should not represent a significant proportion of the cost of the storage system, particularly given the expense of high-capacity SSD devices of the type seen in modern systems.

Crucially, many of these technologies are bistable meaning that patterns on the display remain indefinitely (subject to a certain level of degradation) after the electric field across the electrophoretic material is cut-off. The last image recorded on the screen will persist without any input power. Displays are non-volatile in that power is not required to maintain a display image, only to change it.

The flexible, compact and bistable nature of the material makes it perfect for use in storage systems for providing fault indicating displays which are still visible after a faulty component in a carrier has been disconnected from the rest of the system.

In an embodiment, the apparatus comprises a plurality of electrophoretic displays that are independently controlled. The polarity of the electric field across each display element can be switched irrespective of the electric field across any of the other elements and thus the appearance of the film or screen can differ from one element to the next. The elements can be arranged to produce symbols or text indicative of the type of fault.

In an embodiment, the storage system comprises a plurality of components and each electrophoretic display is associated with one of these components. Each element is configured to indicate whether or not a particular component of the storage system is faulty or is functioning well. A single element will have a substantially uniform appearance since the polarity of the field that determines what colour substance is visible should not vary within an element itself. A one to one correspondence between display elements and components is efficient in terms of both space and cost and will be sufficient if all that is required is to identify faulty components. However, if it is desired to display further information, regarding the type of fault for example, it may be necessary to have several display elements associated with each component. This way the set of display elements can form a pixelated screen and it is possible to present text and/or images.

In an embodiment, the apparatus is not visible when the components are in use.

Generally, it is preferable to mount the display on the carrier itself near to the components to which the display relates. The majority of the surface area of a carrier is not visible when it is installed and connected. Consequently, a display mounted on the carrier will generally also not be visible when the carlier is connected. This positioning is convenient and, since an external LED can be included in the system to inform the user of a fault within a carrier, viewing the display during use is not really necessary.

In an embodiment, the electrophoretic displays are arranged adjacent to one another providing a compact configuration which can be easily located and contains information on the status of plural components in a carrier.

In an embodiment, the electrophoretic displays are controlled so as to display alphanumeric data. Rather than simply indicating with a particular colour whether or not a component is faulty, by placing displays next to each other to form a screen (each element representing a pixel on the screen) it is also possible to form words, pictures or symbols which can provide more information regarding any faults (for example the position of a faulty component or error codes for each fault).

In an embodiment, the electrophoretic displays are arranged so that their relative positions correspond to the relative positions of the associated components. This is a simple way to indicate a fault which will make it easy for a user to quickly locate the relevant component. Few displays are required and these can be mounted together at a convenient location on the carrier.

In an embodiment, the electrophoretic displays are mounted adjacent to the associated component. This makes it easy for a user to ascertain both which component is faulty and where on the carrier this component is located. Likewise if it is desired to check the status of one component in particular this can be easily achieved.

In an embodiment, the fault information is updated at regular intervals. When a particular component fails, in order for this to register on the display the fault must be detected and a message sent to the display or displays associated with the component to switch the polarity of the electric field on that display or displays. This will cause its appearance of the display to change to indicate a fault on the associated component.

Obviously if a number of displays make up the display for one component, field polarity may remain unchanged on some of the displays. Checks can be automated to occur at regular intervals. The user does not need to remember to update the display screen before removal of the carrier housing a faulty module.

In an embodiment, the fault information is updated in response to a user request.

Updating once in response to user input is simple and if the user inputs his/her request as soon as a faulty component is detected then this reduces the possibility of a user forgetting to update the display before removing the carrier from the enclosure.

In an embodiment, the storage system comprises a system controller and fault information is updated in response to detection of a fault by the system controller. This method is efficient and the display will always be updated to indicate a faulty component before removal of the carrier.

According to a second aspect of the present invention, there is provided a method for indicating fault information relating to a component of a data storage system, the method comprising: providing a storage system comprising: a plurality of storage media; and an apparatus having a display arranged to display the fault information; updating the display, wherein the display is configured such that when power to the apparatus is cut-off the fault information remains displayed.

According to another aspect of the present invention, there is provided an apparatus for indicating fault information relating to a component of a data storage system, the apparatus comprising a display arranged to display the fault information and configured such that when power to the apparatus is cut-off the fault information remains displayed.

Embodiments of the present invention will now be described in detail with reference to the accompanying drawings, in which: Figure 1 shows a drive disk enclosure housing a number of vertically orientated carriers; Figure 2 shows the drive disk enclosure of figure 1 with one carrier disconnected and extracted; Figure 3 shows a close up of the extracted carrier of figure 2; Figure 4 shows a plan view of the extracted carrier of figures 2 and 3; Figure 5 shows a close up view of the display portion of an apparatus according to the present invention; Figure 6 shows a storage enclosure with drives mounted in trays. One tray is shown disconnected and extracted from the system; Figure 7 shows the extracted tray of figure 6 with displays mounted on the rear; Figure 8 shows a close up of the displays of figure 7; Figure 9 shows an application server module with the lid removed for servicing; Figure 10 shows a close up of a number of non-volatile displays mounted on the application server module of figure 9, configured to show human readable fault codes and diagnosis tools.

Figure 2 shows an enclosure 4 from a storage system designed by the present applicant, for mounting in a rack. Carriers 6, each holding one or more solid state drives 16 for data storage, are mounted vertically along the front side 10 of the enclosure. One of the SSD5 16 in carrier 12 has failed and the carrier has been disconnected from the midplane (not shown) and extracted from the enclosure. Fans and electronics modules

S

are usually housed in the back portion S of the enclosure and are also connected to the midplane. A close-up view of the carrier 12 is shown in figures 3 and 4.

In the event of a fault on carrier 12, a conventional LED fault indicator will light up or change colour indicating to a user which carrier contains a faulty SSD. LEDS may be mounted on the midplane or another internal location and relayed by a passive lightpipe to enable them to be viewed externally or they may be mounted so that the LED itself is directly visible. At this point, while the carrier is still connected, the electrophoretic display can be updated in order to indicate which of the SSD5 within the carrier is faulty.

An electrophoretic display portion (or strip) 18 is shown clearly in figure 4 (close up in figure 5) fixed to the inner surface of the carrier adjacent to the drives 16 and other components present. Fixing can be achieved, for example, by means of a small clip or using adhesive and may be fixed temporarily or permanently to the carrier. An ERG (flexible printed circuit) connector 20 carries control signals for the display and can be used as the clip or fixing means. The display portion 18 is divided into several displays 22 (shown in figure 5). Each one of these displays corresponds to one of the SSD5 present in the carrier. Each display can be independently controlled by inducing an electric field across it and changing the polarity of the electrodes to change the colour of the uppermost side of the display. In figure 5, black displays indicate faulty or absent SSDs and white displays indicate fully functioning SSD5. Of course, colours can be reversed and need not necessarily be black and white but can be any colour so long as when a fault is indicated, the display is clearly distinguishable from its state when the corresponding component is not faulty. The circuitry required to alter the orientation of the electric field across regions of the electrophoretic material corresponding to faulty drives will be evident to one skilled in the ad and so will not be covered in detail here.

Once the display portion has been updated, the carrier is removed from the enclosure.

Due to the bistable properties of the electrophoretic material used, once power to the drives has been disconnected the displays indicating faulty or absent drives will remain black and those indicating working drives will remain white (as shown in figure 5). This is in contrast to prior art solutions, none of which can easily indicate a fault on a carrier which has been removed from the system. Thus it is possible for a user to leave the carrier for a time before repairing or to pass it to another user for repair at which point it will still be clearly indicated which component or components require replacement.

In order to achieve a working display after removal of a carrier from an enclosure, prior art methods would have used complex, bulky circuitry which would have needed to include an additional power source (a battery or supercapacitor), latched data, a push button and LED5. The LED5 could be configured to light up in dependence on the state of the latched data when the button was pushed. Added to other obvious size, time and cost drawbacks to this approach is the fad that the power supply will eventually fail at which time it will no longer be possible to identify faulty components.

The electrophoretic displays on the display portion can be arranged to reflect the position of the components with which they are associated making locating faulty components easier. Figures 6 and 7 illustrate a further embodiment in which this feature has been incorporated. In figure 6 a storage system able to hold eight trays 24 mounted on racks is shown. Tray 26 has been removed from the rack and drives 28 are shown mounted horizontally within the tray. The circle in figure 7 indicates the position of the electrophoretic display portion 30 on the printed circuit board assembly 32 at the rear of the tray 26. A close-up view of the display portion is shown in figure 8. The positioning of the displays (or pixels) 34, each of which can be controlled separately similarly to the display strip of the previous embodiment, corresponds to the physical layout of the disks within the tray. For example, display X on the display portion shown in figure 8 corresponds to disk X in the tray shown in figure 7 and display Y in figure 8 to disk Y in figure 7. A fault within a particular tray will be indicated by a conventional LED indicator mounted externally, at which point the electrophoretic display portion is updated so that faulty or absent drives within the tray are indicated. The tray can then be removed from the rack and the display portion examined. If display X appears black and display Y appears white then the user knows that drive X is faulty and drive Y is functional.

Alternatively, displays (or display portions formed of several displays) can be located next to the component to which they relate.

Figures 9 and 10 illustrate an embodiment wherein the electrophoretic display portion 38 consists of small displays arranged in order to represent pixels on a screen. The resulting image is human readable, i.e. is configured to display words, numbers or symbols to indicate which components are faulty and to assist a user in diagnosing the fault. Figure 9 illustrates an application server module 36 with the lid removed for servicing and the electrophoretic display portion 38 visible on the upper surface. Figure shows a more detailed view of the display portion in which fault codes and diagnostic aids are visible.

Embodiments of the present invention have been described with particular reference to the examples illustrated. However, it will be appreciated that variations and modifications may be made to the examples described within the scope of the present invention.

Claims

Claims 1. A storage system, comprising: a plurality of storage media; and an apparatus for indicating fault information relating to a component of the storage system, the apparatus comprising a display arranged to display the fault information and configured such that when power to the apparatus is cut-off the fault information remains displayed.
2. The storage system of claim 1, wherein the apparatus is mounted on a carrier that can be disconnected and removed from the storage system.
3. The storage system of any of claims 1 and 2, wherein the display is an electrophoretic display.
4. The storage system of claim 3 wherein the apparatus comprises a plurality of electrophoretic displays that are independently controlled.
5. The storage system of claim 4, wherein the storage system comprises a plurality of components and each of the plurality of electrophoretic displays is associated with one of these components.
6. The storage system of claim 5, wherein the apparatus is not visible when the components are in use.
7. The storage system of any of claims 4 to 6, wherein the electrophoretic displays are arranged adjacent to one another.
8. The storage system of claim 7, wherein the electrophoretic displays are controlled so as to display alphanumeric data.
9. The storage system of any of claims 4 to 6, wherein the electrophoretic displays are arranged so that their relative positions correspond to the relative positions of the associated components.
10. The storage system of any of claims 4 to 6, wherein the electrophoretic displays are mounted adjacent to the associated component.
11. The storage system of any of claims ito 10, wherein the fault information is updated at regular intervals.
12. The storage system of any of claims ito 10, wherein the fault information is updated in response to a user request.
13. The storage system of any of claims 1 to 10, wherein the storage system comprises a system controller and fault information is updated in response to detection of a fault by the system controller.
14. A method for indicating fault information relating to a component of a data storage system, the method comprising: providing a storage system comprising: a plurality of storage media; and an apparatus having a display arranged to display the fault information; updating the display, wherein the display is configured such that when power to the apparatus is cut-off the fault information remains displayed.
15. A storage system comprising a plurality of storage media and an apparatus for displaying fault information relating to a component of the storage system, the apparatus being substantially as shown in and/or described with reference to any one or more of Figures ito 10 of the accompanying drawings.
16. A method for displaying fault information relating to a component of a data storage system, the method being substantially as shown in and/or described with reference to any one or more of Figures 1 to 10 of the accompanying drawings.